bac-analyzer
Pass
Audited by Gen Agent Trust Hub on Nov 9, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTIONDATA_EXFILTRATION
Full Analysis
- [SAFE]: The skill's functionality and implementation are consistent with its described role as a passive security analysis tool. No malicious code, obfuscation, or unauthorized data access was detected.
- [INDIRECT_PROMPT_INJECTION]: The skill analyzes untrusted data from external traffic captures, which serves as a potential surface for indirect prompt injection if the output is processed by the agent without verification.
- Ingestion points:
scripts/parse-har.py,scripts/parse-burp.py, andscripts/parse-caido.pyingest data from external files and the Caido API. - Boundary markers: The scripts do not implement explicit boundary markers in the generated reports or console output.
- Capability inventory: The skill can execute local scripts and make network requests to a user-specified Caido instance.
- Sanitization: Data ingestion relies on standard parsers (
json,xml.etree.ElementTree) to ensure structural integrity. - [COMMAND_EXECUTION]: The orchestrator script
scripts/analyze.shmanages the analysis workflow by executing multiple internal Python scripts and handling temporary file operations. - [DATA_EXFILTRATION]: The skill facilitates the ingestion of traffic data from a Caido API as a documented feature.
- Evidence:
scripts/parse-caido.pyconnects to the URL provided in theCAIDO_URLenvironment variable and uses theCAIDO_API_KEYfor authentication to that specific instance.
Audit Metadata